Re: [xml] question about getting numbers out of an xml file

[Daniel Veillard <veillard redhat com>]

On Sun, Mar 18, 2007 at 12:55:14AM -0400, Stefan Jeglinski wrote:
> I've been looking through the archives, and reading the API, and I'm 
> curious about something that must be a simple thing.
> 
> I am using a small xml file to store some numbers (ints and floats). 
> I am trying to use an xpath to extract them.
> 
> So, I started this way (sort of pseudo C code):
> 
> doc      = xmlParseFile(filename);
> xpathCtx = xmlXPathNewContext(doc);
> xpath    = "/path/to/element";
> xpathObj = xmlXPathEvalExpression(xpathExpr, xpathCtx);
> content  = xmlNodeGetContent(xpathObj->nodesetval->nodeTab[0]);
> 
> This gives me the content of the element as an xmlChar* (it works - 
> if it shouldn't then I made a typo). Of course since the element 
> content is to be used as a number (int, float) then I have to further 
> convert it from a string to a number.
> 
> So, looking for something more elegant, I came across the function:
> 
> 	xmlXPathCastToNumber(xpathObj)
> 
> This works fine returning a double. But I also came across the function:
> 
> 	xmlXPathStringEvalNumber(xpathObj->stringval);
> 		or
> 	xmlXPathStringEvalNumber(xpathObj->floatval);

  Wrong approach ... using the semi-internals functions of XPath module
to do the work instead of using XPath to do the extraction in a portable
way is really not 'elegant' from my point of view
  see 
    http://www.w3.org/TR/xpath#numbers
  and
    http://www.w3.org/TR/xpath#function-number

  something like xpath="number(/path/to/element)"
The evaluation will bring back a number , i.e. the xpathObj will be 
different but you will get the float directly.

Daniel

-- 
Red Hat Virtualization group http://redhat.com/virtualization/
Daniel Veillard      | virtualization library  http://libvirt.org/
veillard redhat com  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine  http://rpmfind.net/